Mineral Replacements in Flooding Experiments Linked to Enhanced Oil Recovery in Chalk

Abstract

Seawater injection into chalk-reservoirs on the Norwegian Continental Shelf has increased the oil recovery and reduced seabed subsidence. In the researchfield of Improved Oil Recovery (IOR) the effects of injection of seawater-like brines into chalk have been studied for decades. Particular ions in brines seem to have the power to change the wetting conditions of the rock as well as affecting the strength of chalk. Therefore, when optimizing brines to enhance the production of oil, it is paramount to understand how the injected brine impose alterations in geo-mechanical properties, which in turn affect the strength of the rock. These effects are referred to as “water weakening of chalk” and should be controlled to avoid undesired compaction effects and loss in well stabilities, affecting both safety and costs. This thesis aims to describe the “the which, the how and the where” of mineralogical alterations in chalk when flooded with reactive brines, with special focus on the Mg2+-ion. The produced changes may directly impact mechanical properties of the rock, and can often only be observed at micron and sub-micron-scale. The study of these alterations therefore requires methods with possibilities to image and quantify the chemical composition of the rock with resolution below pore-scale, often at nano-scale. The analytical tools used in this study are, on one hand, well-known in terms of their application to rocks and partly on chalk, but on the other hand, this research has developed new methodological approaches for the study of mineralogical changes in Enhanced Oil Recovery (EOR) experiments. This work has resulted in the design of a toolbox which holds the possibility to sufficiently investigate the mineralogical effects of EOR-fluids. Tests have been performed on several types of outcrop chalk, flooded with different brines, at different stresses and temperatures, and for different periods of time. In addition, similar experiments on “artificial chalk cores”, made from pure calcite powder, have been completed. The tests were performed on cylindrical core samples and analysed by scanning and transmission electron microscopy, Mineral Liberation Analyzer, electron microprobe analysis, whole-rock geochemistry, stable isotope analyses, nano secondary ion mass spectrometry, X-ray diffraction, along with measurements of the specific surface area, density and porosity, and quantification of the composition of the effluent water. Thorough core- pore- and nano-scale investigation has been conducted and the results from all scales match. The combination of scales supplement each other to an improved understanding of the data and the following results can be highlighted: Analyses after flooding with NaCl show very few mineralogical reactions in chalk, however, in silica-rich chalk, the distribution of silicate minerals may be significantly altered. After flooding with MgCl2, the results show dissolution of calcite and precipitation of magnesite. The occurrence and shape of new-grown crystals depend on flooding time and distance from the flooding inlet of the sample cores, together with the primary mineralogy of the chalk and its diagenetic history. Crystals vary in size, from a few nanometres up to over 10 μm and may crystallize as single grains or in clusters. Additionally, noncarbonate phases dissolve and precipitate during flooding, altering the distribution of these minerals within the cores. The type of chalk, with different contents and types of non-carbonate minerals, is found to play a role for the strength of the chalk and is also reflected in the precipitated mineral phases during flooding. These new-formed minerals may alter the permeability, porosity and the reactive surface of the flooded chalk. All tests show pronounced alteration of texture and mineralogy at the flow-inlet side of the core, along with a decreasing trend in magnesium content towards the outlet. With longer duration of flooding, the alterations move like a front further into the cores, and for a three year long test, the whole core is altered from the primary mineralogy to newly formed minerals. When studied at core- and pore-scale, the newly formed crystals are found to be magnesite with minor calcium impurities, together with clay minerals. In two slightly shorter tests, flooded for one and a half and two years, respectively, the alteration front is still observable. On the outlet side of the cores, the mineralogy still mainly consists of calcite, primary clay minerals and other non-carbonate minerals together with occurrences of newly formed magnesite and secondary clay minerals. On the inlet side of the cores, the mineralogy consists of magnesite and clay minerals, as observed in the experiment flooded for three years. Dolomite or low to high Mg-calcite, are not observed. An interesting observation in long-term flooded chalk, is the abrupt transition between the two mineral regimes found on each side of the alteration front. These transition zones have higher porosity than other parts of the cores, a pattern similar to what is observed in single-crystal experiments, where the alterations happen through dissolution and precipitation, driven by the progression of high porosity-zones and the state of equilibrium at the boundary between the primary and secondary mineral phases. In long-term flooded chalk, the texture of larger macrofossils is often preserved, while their mineralogy is altered. Such pseudomorphism is observed at centimetre- and micrometre-scale, but not at nanometre-scale, pointing to the precipitation rate of magnesite being the rate-limiting factor, found to result in micrometre-scale pseudomorphism. Severe signs of dissolution on calcite grains and high-resolution analyses of precipitation on grain-scale confirm that the formation of new mineral phases is controlled by dissolution and precipitation, and not by molecular solid state diffusion. [...]